Andrea Soddu

Diagnostic precision of PET imaging and functional MRI in disorders of consciousness: a clinical validation study

BACKGROUNDnBedside clinical examinations can have high rates of misdiagnosis of unresponsive wakefulness syndrome (vegetative state) or minimally conscious state. The diagnostic and prognostic usefulness of neuroimaging-based approaches has not been established in a clinical setting. We did a validation study of two neuroimaging-based diagnostic methods: PET imaging and functional MRI (fMRI).nnnMETHODSnFor this clinical validation study, we included patients referred to the University Hospital of Liège, Belgium, between January, 2008, and June, 2012, who were diagnosed by our unit with unresponsive wakefulness syndrome, locked-in syndrome, or minimally conscious state with traumatic or non-traumatic causes. We did repeated standardised clinical assessments with the Coma Recovery Scale-Revised (CRS-R), cerebral (18)F-fluorodeoxyglucose (FDG) PET, and fMRI during mental activation tasks. We calculated the diagnostic accuracy of both imaging methods with CRS-R diagnosis as reference. We assessed outcome after 12 months with the Glasgow Outcome Scale-Extended.nnnFINDINGSnWe included 41 patients with unresponsive wakefulness syndrome, four with locked-in syndrome, and 81 in a minimally conscious state (48=traumatic, 78=non-traumatic; 110=chronic, 16=subacute). (18)F-FDG PET had high sensitivity for identification of patients in a minimally conscious state (93%, 95% CI 85-98) and high congruence (85%, 77-90) with behavioural CRS-R scores. The active fMRI method was less sensitive at diagnosis of a minimally conscious state (45%, 30-61) and had lower overall congruence with behavioural scores (63%, 51-73) than PET imaging. (18)F-FDG PET correctly predicted outcome in 75 of 102 patients (74%, 64-81), and fMRI in 36 of 65 patients (56%, 43-67). 13 of 41 (32%) of the behaviourally unresponsive patients (ie, diagnosed as unresponsive with CRS-R) showed brain activity compatible with (minimal) consciousness (ie, activity associated with consciousness, but diminished compared with fully conscious individuals) on at least one neuroimaging test; 69% of these (9 of 13) patients subsequently recovered consciousness.nnnINTERPRETATIONnCerebral (18)F-FDG PET could be used to complement bedside examinations and predict long-term recovery of patients with unresponsive wakefulness syndrome. Active fMRI might also be useful for differential diagnosis, but seems to be less accurate.nnnFUNDINGnThe Belgian National Funds for Scientific Research (FNRS), Fonds Léon Fredericq, the European Commission, the James McDonnell Foundation, the Mind Science Foundation, the French Speaking Community Concerted Research Action, the University of Copenhagen, and the University of Liège.

Intrinsic functional connectivity differentiates minimally conscious from unresponsive patients

Despite advances in resting state functional magnetic resonance imaging investigations, clinicians remain with the challenge of how to implement this paradigm on an individualized basis. Here, we assessed the clinical relevance of resting state functional magnetic resonance imaging acquisitions in patients with disorders of consciousness by means of a systems-level approach. Three clinical centres collected data from 73 patients in minimally conscious state, vegetative state/unresponsive wakefulness syndrome and coma. The main analysis was performed on the data set coming from one centre (Liège) including 51 patients (26 minimally conscious state, 19 vegetative state/unresponsive wakefulness syndrome, six coma; 15 females; mean age 49 ± 18 years, range 11-87; 16 traumatic, 32 non-traumatic of which 13 anoxic, three mixed; 35 patients assessed >1 month post-insult) for whom the clinical diagnosis with the Coma Recovery Scale-Revised was congruent with positron emission tomography scanning. Group-level functional connectivity was investigated for the default mode, frontoparietal, salience, auditory, sensorimotor and visual networks using a multiple-seed correlation approach. Between-group inferential statistics and machine learning were used to identify each networks capacity to discriminate between patients in minimally conscious state and vegetative state/unresponsive wakefulness syndrome. Data collected from 22 patients scanned in two other centres (Salzburg: 10 minimally conscious state, five vegetative state/unresponsive wakefulness syndrome; New York: five minimally conscious state, one vegetative state/unresponsive wakefulness syndrome, one emerged from minimally conscious state) were used to validate the classification with the selected features. Coma Recovery Scale-Revised total scores correlated with key regions of each network reflecting their involvement in consciousness-related processes. All networks had a high discriminative capacity (>80%) for separating patients in a minimally conscious state and vegetative state/unresponsive wakefulness syndrome. Among them, the auditory network was ranked the most highly. The regions of the auditory network which were more functionally connected in patients in minimally conscious state compared to vegetative state/unresponsive wakefulness syndrome encompassed bilateral auditory and visual cortices. Connectivity values in these three regions discriminated congruently 20 of 22 independently assessed patients. Our findings point to the significance of preserved abilities for multisensory integration and top-down processing in minimal consciousness seemingly supported by auditory-visual crossmodal connectivity, and promote the clinical utility of the resting paradigm for single-patient diagnostics.

Hypnotic modulation of resting state fMRI default mode and extrinsic network connectivity

Resting state fMRI (functional magnetic resonance imaging) acquisitions are characterized by low-frequency spontaneous activity in a default mode network (encompassing medial brain areas and linked to self-related processes) and an anticorrelated extrinsic system (encompassing lateral frontoparietal areas and modulated via external sensory stimulation). In order to better determine the functional contribution of these networks to conscious awareness, we here sought to transiently modulate their relationship by means of hypnosis. We used independent component analysis (ICA) on resting state fMRI acquisitions during normal wakefulness, under hypnotic state, and during a control condition of autobiographical mental imagery. As compared to mental imagery, hypnosis-induced modulation of resting state fMRI networks resulted in a reduced extrinsic lateral frontoparietal cortical connectivity, possibly reflecting a decreased sensory awareness. The default mode network showed an increased connectivity in bilateral angular and middle frontal gyri, whereas its posterior midline and parahippocampal structures decreased their connectivity during hypnosis, supposedly related to an altered self awareness and posthypnotic amnesia. In our view, fMRI resting state studies of physiological (e.g., sleep or hypnosis), pharmacological (e.g., sedation or anesthesia), and pathological modulation (e.g., coma or related states) of intrinsic default mode and anticorrelated extrinsic sensory networks, and their interaction with other cerebral networks, will further improve our understanding of the neural correlates of subjective awareness.

Biased binomial assessment of cross-validated estimation of classification accuracies illustrated in diagnosis predictions

Multivariate classification is used in neuroimaging studies to infer brain activation or in medical applications to infer diagnosis. Their results are often assessed through either a binomial or a permutation test. Here, we simulated classification results of generated random data to assess the influence of the cross-validation scheme on the significance of results. Distributions built from classification of random data with cross-validation did not follow the binomial distribution. The binomial test is therefore not adapted. On the contrary, the permutation test was unaffected by the cross-validation scheme. The influence of the cross-validation was further illustrated on real-data from a brain–computer interface experiment in patients with disorders of consciousness and from an fMRI study on patients with Parkinson disease. Three out of 16 patients with disorders of consciousness had significant accuracy on binomial testing, but only one showed significant accuracy using permutation testing. In the fMRI experiment, the mental imagery of gait could discriminate significantly between idiopathic Parkinsons disease patients and healthy subjects according to the permutation test but not according to the binomial test. Hence, binomial testing could lead to biased estimation of significance and false positive or negative results. In our view, permutation testing is thus recommended for clinical application of classification with cross-validation.

An independent SSVEP-based brain-computer interface in locked-in syndrome

OBJECTIVEnSteady-state visually evoked potential (SSVEP)-based brain-computer interfaces (BCIs) allow healthy subjects to communicate. However, their dependence on gaze control prevents their use with severely disabled patients. Gaze-independent SSVEP-BCIs have been designed but have shown a drop in accuracy and have not been tested in brain-injured patients. In the present paper, we propose a novel independent SSVEP-BCI based on covert attention with an improved classification rate. We study the influence of feature extraction algorithms and the number of harmonics. Finally, we test online communication on healthy volunteers and patients with locked-in syndrome (LIS).nnnAPPROACHnTwenty-four healthy subjects and six LIS patients participated in this study. An independent covert two-class SSVEP paradigm was used with a newly developed portable light emitting diode-based interlaced squares stimulation pattern.nnnMAIN RESULTSnMean offline and online accuracies on healthy subjects were respectively 85 ± 2% and 74 ± 13%, with eight out of twelve subjects succeeding to communicate efficiently with 80 ± 9% accuracy. Two out of six LIS patients reached an offline accuracy above the chance level, illustrating a response to a command. One out of four LIS patients could communicate online.nnnSIGNIFICANCEnWe have demonstrated the feasibility of online communication with a covert SSVEP paradigm that is truly independent of all neuromuscular functions. The potential clinical use of the presented BCI system as a diagnostic (i.e., detecting command-following) and communication tool for severely brain-injured patients will need to be further explored.

Neural correlates of consciousness in patients who have emerged from a minimally conscious state: a cross-sectional multimodal imaging study

BACKGROUNDnBetween pathologically impaired consciousness and normal consciousness exists a scarcely researched transition zone, referred to as emergence from minimally conscious state, in which patients regain the capacity for functional communication, object use, or both. We investigated neural correlates of consciousness in these patients compared with patients with disorders of consciousness and healthy controls, by multimodal imaging.nnnMETHODSnIn this cross-sectional, multimodal imaging study, patients with unresponsive wakefulness syndrome, patients in a minimally conscious state, and patients who had emerged from a minimally conscious state, diagnosed with the Coma Recovery Scale-Revised, were recruited from the neurology department of the Centre Hospitalier Universitaire de Liège, Belgium. Key exclusion criteria were neuroimaging examination in an acute state, sedation or anaesthesia during scanning, large focal brain damage, motion parameters of more than 3 mm in translation and 3° in rotation, and suboptimal segmentation and normalisation. We acquired resting state functional and structural MRI data and (18)F-fluorodeoxyglucose (FDG) PET data; we used seed-based functional MRI (fMRI) analysis to investigate positive default mode network connectivity (within-network correlations) and negative default mode network connectivity (between-network anticorrelations). We correlated FDG-PET brain metabolism with fMRI connectivity. We used voxel-based morphometry to test the effect of anatomical deformations on functional connectivity.nnnFINDINGSnWe recruited a convenience sample of 58 patients (21 [36%] with unresponsive wakefulness syndrome, 24 [41%] in a minimally conscious state, and 13 [22%] who had emerged from a minimally conscious state) and 35 healthy controls between Oct 1, 2009, and Oct 31, 2014. We detected consciousness-level-dependent increases (from unresponsive wakefulness syndrome, minimally conscious state, emergence from minimally conscious state, to healthy controls) for positive and negative default mode network connectivity, brain metabolism, and grey matter volume (p<0·05 false discovery rate corrected for multiple comparisons). Positive default mode network connectivity differed between patients and controls but not among patient groups (F test p<0·0001). Negative default mode network connectivity was only detected in healthy controls and in those who had emerged from a minimally conscious state; patients with unresponsive wakefulness syndrome or in a minimally conscious state showed pathological between-network positive connectivity (hyperconnectivity; F test p<0·0001). Brain metabolism correlated with positive default mode network connectivity (Spearmans r=0·50 [95% CI 0·26 to 0·61]; p<0·0001) and negative default mode network connectivity (Spearmans r=-0·52 [-0·35 to -0·67); p<0·0001). Grey matter volume did not differ between the studied groups (F test p=0·06).nnnINTERPRETATIONnPartial preservation of between-network anticorrelations, which are seemingly of neuronal origin and cannot be solely explained by morphological deformations, characterise patients who have emerged from a minimally conscious state. Conversely, patients with disorders of consciousness show pathological between-network correlations. Apart from a deeper understanding of the neural correlates of consciousness, these findings have clinical implications and might be particularly relevant for outcome prediction and could inspire new therapeutic options.nnnFUNDINGnBelgian National Funds for Scientific Research (FNRS), European Commission, Natural Sciences and Engineering Research Council of Canada, James McDonnell Foundation, European Space Agency, Mind Science Foundation, French Speaking Community Concerted Research Action, Fondazione Europea di Ricerca Biomedica, University and University Hospital of Liège (Liège, Belgium), and University of Western Ontario (London, ON, Canada).

Lies, damned lies and diagnoses: estimating the clinical utility of assessments of covert awareness in the vegetative state.

Abstract Background: Functional neuroimaging of patients in the vegetative state has been shown to provide diagnostic and prognostic information beyond that which conventional behavioural assessments may allow. However, before these promising approaches may reach large numbers of patients through a standard clinical protocol, it is necessary to determine the utility of these assessments—i.e. the accuracy of their diagnoses. Methods and results: This study demonstrated that, due to the nature of statistical testing and the absence of a ‘ground truth’ of consciousness, it is impossible to calculate the conventional measures of clinical utility—sensitivity and specificity—for diagnoses made on the basis of functional neuroimaging for command-following. Nevertheless, it is crucial for such measures to be determined in order for valuable clinical resources to be distributed wisely. Therefore, a number of alternative guidelines are offered for the estimation of clinical utility. Conclusions: By evaluating new and existing functional neuroimaging methods against the proposed guidelines, this study argues that it may be possible to achieve dramatically and efficiently improved diagnostic and prognostic accuracy for all vegetative state patients.

Cerebral functional connectivity periodically (de)synchronizes with anatomical constraints

This paper studies the link between resting-state functional connectivity (FC), measured by the correlations of fMRI BOLD time courses, and structural connectivity (SC), estimated through fiber tractography. Instead of a static analysis based on the correlation between SC and FC averaged over the entire fMRI time series, we propose a dynamic analysis, based on the time evolution of the correlation between SC and a suitably windowed FC. Assessing the statistical significance of the time series against random phase permutations, our data show a pronounced peak of significance for time window widths around 20–30 TR (40–60 s). Using the appropriate window width, we show that FC patterns oscillate between phases of high modularity, primarily shaped by anatomy, and phases of low modularity, primarily shaped by inter-network connectivity. Building upon recent results in dynamic FC, this emphasizes the potential role of SC as a transitory architecture between different highly connected resting-state FC patterns. Finally, we show that the regions contributing the most to these whole-brain level fluctuations of FC on the supporting anatomical architecture belong to the default mode and the executive control networks suggesting that they could be capturing consciousness-related processes such as mind wandering.

Correlation between resting state fMRI total neuronal activity and PET metabolism in healthy controls and patients with disorders of consciousness

The mildly invasive 18F‐fluorodeoxyglucose positron emission tomography (FDG‐PET) is a well‐established imaging technique to measure ‘resting state’ cerebral metabolism. This technique made it possible to assess changes in metabolic activity in clinical applications, such as the study of severe brain injury and disorders of consciousness.

Highlighting the Structure-Function Relationship of the Brain with the Ising Model and Graph Theory

With the advent of neuroimaging techniques, it becomes feasible to explore the structure-function relationships in the brain. When the brain is not involved in any cognitive task or stimulated by any external output, it preserves important activities which follow well-defined spatial distribution patterns. Understanding the self-organization of the brain from its anatomical structure, it has been recently suggested to model the observed functional pattern from the structure of white matter fiber bundles. Different models which study synchronization (e.g., the Kuramoto model) or global dynamics (e.g., the Ising model) have shown success in capturing fundamental properties of the brain. In particular, these models can explain the competition between modularity and specialization and the need for integration in the brain. Graphing the functional and structural brain organization supports the model and can also highlight the strategy used to process and organize large amount of information traveling between the different modules. How the flow of information can be prevented or partially destroyed in pathological states, like in severe brain injured patients with disorders of consciousness or by pharmacological induction like in anaesthesia, will also help us to better understand how global or integrated behavior can emerge from local and modular interactions.